/* xbt_str.c - various helping functions to deal with strings */
-/* Copyright (c) 2007, 2008, 2009, 2010. The SimGrid Team.
+/* Copyright (c) 2007-2014. The SimGrid Team.
* All rights reserved. */
/* This program is free software; you can redistribute it and/or modify it
* under the terms of the license (GNU LGPL) which comes with this package. */
-#include "portable.h"
+#include "src/portable.h"
#include "xbt/misc.h"
#include "xbt/sysdep.h"
#include "xbt/str.h" /* headers of these functions */
* Strips the whitespaces from the end of s.
* By default (when char_list=NULL), these characters get stripped:
*
- * - " " (ASCII 32 (0x20)) space.
- * - "\t" (ASCII 9 (0x09)) tab.
- * - "\n" (ASCII 10 (0x0A)) line feed.
- * - "\r" (ASCII 13 (0x0D)) carriage return.
- * - "\0" (ASCII 0 (0x00)) NULL.
- * - "\x0B" (ASCII 11 (0x0B)) vertical tab.
+ * - " " (ASCII 32 (0x20)) space.
+ * - "\t" (ASCII 9 (0x09)) tab.
+ * - "\n" (ASCII 10 (0x0A)) line feed.
+ * - "\r" (ASCII 13 (0x0D)) carriage return.
+ * - "\0" (ASCII 0 (0x00)) NULL.
+ * - "\x0B" (ASCII 11 (0x0B)) vertical tab.
*
* @param s The string to strip. Modified in place.
* @param char_list A string which contains the characters you want to strip.
* Strips the whitespaces from the begining of s.
* By default (when char_list=NULL), these characters get stripped:
*
- * - " " (ASCII 32 (0x20)) space.
- * - "\t" (ASCII 9 (0x09)) tab.
- * - "\n" (ASCII 10 (0x0A)) line feed.
- * - "\r" (ASCII 13 (0x0D)) carriage return.
- * - "\0" (ASCII 0 (0x00)) NULL.
- * - "\x0B" (ASCII 11 (0x0B)) vertical tab.
+ * - " " (ASCII 32 (0x20)) space.
+ * - "\t" (ASCII 9 (0x09)) tab.
+ * - "\n" (ASCII 10 (0x0A)) line feed.
+ * - "\r" (ASCII 13 (0x0D)) carriage return.
+ * - "\0" (ASCII 0 (0x00)) NULL.
+ * - "\x0B" (ASCII 11 (0x0B)) vertical tab.
*
* @param s The string to strip. Modified in place.
* @param char_list A string which contains the characters you want to strip.
* Strips the whitespaces from both the beginning and the end of s.
* By default (when char_list=NULL), these characters get stripped:
*
- * - " " (ASCII 32 (0x20)) space.
- * - "\t" (ASCII 9 (0x09)) tab.
- * - "\n" (ASCII 10 (0x0A)) line feed.
- * - "\r" (ASCII 13 (0x0D)) carriage return.
- * - "\0" (ASCII 0 (0x00)) NULL.
- * - "\x0B" (ASCII 11 (0x0B)) vertical tab.
+ * - " " (ASCII 32 (0x20)) space.
+ * - "\t" (ASCII 9 (0x09)) tab.
+ * - "\n" (ASCII 10 (0x0A)) line feed.
+ * - "\r" (ASCII 13 (0x0D)) carriage return.
+ * - "\0" (ASCII 0 (0x00)) NULL.
+ * - "\x0B" (ASCII 11 (0x0B)) vertical tab.
*
* @param s The string to strip.
* @param char_list A string which contains the characters you want to strip.
xbt_str_ltrim(s, char_list);
}
-/** @brief Replace double whitespaces (but no other characters) from the string.
- *
- * The function modifies the string so that each time that several spaces appear,
- * they are replaced by a single space. It will only do so for spaces (ASCII 32, 0x20).
- *
- * @param s The string to strip. Modified in place.
- *
- */
-void xbt_str_strip_spaces(char *s)
-{
- char *p = s;
- int e = 0;
-
- if (!s)
- return;
-
- while (1) {
- if (!*p)
- goto end;
-
- if (*p != ' ')
- break;
-
- p++;
- }
-
- e = 1;
-
- do {
- if (e)
- *s++ = *p;
-
- if (!*++p)
- goto end;
-
- if (e ^ (*p != ' '))
- if ((e = !e))
- *s++ = ' ';
- } while (1);
-
-end:
- *s = '\0';
-}
-
/** @brief Substitutes a char for another in a string
*
* @param str the string to modify
*
* By default (with sep=NULL), these characters are used as separator:
*
- * - " " (ASCII 32 (0x20)) space.
- * - "\t" (ASCII 9 (0x09)) tab.
- * - "\n" (ASCII 10 (0x0A)) line feed.
- * - "\r" (ASCII 13 (0x0D)) carriage return.
- * - "\0" (ASCII 0 (0x00)) NULL.
- * - "\x0B" (ASCII 11 (0x0B)) vertical tab.
+ * - " " (ASCII 32 (0x20)) space.
+ * - "\t" (ASCII 9 (0x09)) tab.
+ * - "\n" (ASCII 10 (0x0A)) line feed.
+ * - "\r" (ASCII 13 (0x0D)) carriage return.
+ * - "\0" (ASCII 0 (0x00)) NULL.
+ * - "\x0B" (ASCII 11 (0x0B)) vertical tab.
*/
xbt_dynar_t xbt_str_split(const char *s, const char *sep)
//if substring was not found add the entire string
if (NULL == q) {
v = strlen(p);
- to_push = malloc(v + 1);
+ to_push = xbt_malloc(v + 1);
memcpy(to_push, p, v);
to_push[v] = '\0';
xbt_dynar_push(res, &to_push);
done = 1;
} else {
//get the appearance
- to_push = malloc(q - p + 1);
+ to_push = xbt_malloc(q - p + 1);
memcpy(to_push, p, q - p);
//add string terminator
to_push[q - p] = '\0';
*
* The string passed as argument must be writable (not const)
* The elements of the dynar are just parts of the string passed as argument.
- *
- * To free the structure constructed by this function, free the first element and free the dynar:
- *
- * free(xbt_dynar_get_ptr(dynar,0));
- * xbt_dynar_free(&dynar);
+ * So if you don't store that argument elsewhere, you should free it in addition
+ * to freeing the dynar. This can be done by simply freeing the first argument
+ * of the dynar:
+ * free(xbt_dynar_get_ptr(dynar,0));
+ *
+ * Actually this function puts a bunch of \0 in the memory area you passed as
+ * argument to separate the elements, and pushes the address of each chunk
+ * in the resulting dynar. Yes, that's uneven. Yes, that's gory. But that's efficient.
*/
xbt_dynar_t xbt_str_split_quoted_in_place(char *s) {
xbt_dynar_t res = xbt_dynar_new(sizeof(char *), NULL);
beg = s;
- /* do not trim leading spaces: caller responsability to clean his cruft */
+ /* do not trim leading spaces: caller responsibility to clean his cruft */
end = beg;
while (!done) {
return res;
}
-#if defined(SIMGRID_NEED_GETLINE) || defined(DOXYGEN)
-/** @brief Get a single line from the stream (reimplementation of the GNU getline)
- *
- * This is a redefinition of the GNU getline function, used on platforms where it does not exists.
- *
- * getline() reads an entire line from stream, storing the address of the buffer
- * containing the text into *buf. The buffer is null-terminated and includes
- * the newline character, if one was found.
+/** @brief creates a new string containing what can be read on a fd
*
- * If *buf is NULL, then getline() will allocate a buffer for storing the line,
- * which should be freed by the user program. Alternatively, before calling getline(),
- * *buf can contain a pointer to a malloc()-allocated buffer *n bytes in size. If the buffer
- * is not large enough to hold the line, getline() resizes it with realloc(), updating *buf and *n
- * as necessary. In either case, on a successful call, *buf and *n will be updated to
- * reflect the buffer address and allocated size respectively.
*/
-long getline(char **buf, size_t * n, FILE * stream)
+char *xbt_str_from_file(FILE * file)
{
+ xbt_strbuff_t buff = xbt_strbuff_new();
+ char *res;
+ char bread[1024];
+ memset(bread, 0, 1024);
- size_t i;
- int ch;
-
- if (!*buf) {
- *buf = xbt_malloc(512);
- *n = 512;
- }
-
- if (feof(stream))
- return (ssize_t) - 1;
-
- for (i = 0; (ch = fgetc(stream)) != EOF; i++) {
-
- if (i >= (*n) + 1)
- *buf = xbt_realloc(*buf, *n += 512);
-
- (*buf)[i] = ch;
-
- if ((*buf)[i] == '\n') {
- i++;
- (*buf)[i] = '\0';
- break;
- }
+ while (!feof(file)) {
+ int got = fread(bread, 1, 1023, file);
+ bread[got] = '\0';
+ xbt_strbuff_append(buff, bread);
}
- if (i == *n)
- *buf = xbt_realloc(*buf, *n += 1);
-
- (*buf)[i] = '\0';
-
- return (ssize_t) i;
+ res = buff->data;
+ xbt_strbuff_free_container(buff);
+ return res;
}
-#endif /* HAVE_GETLINE */
-
-/*
- * Diff related functions
+/** @brief Parse an integer out of a string, or raise an error
+ *
+ * The #str is passed as argument to your #error_msg, as follows:
+ * THROWF(arg_error, 0, error_msg, str);
*/
-static XBT_INLINE int diff_get(xbt_matrix_t C, int i, int j)
-{
- return (i == -1 || j == -1) ? 0 : xbt_matrix_get_as(C, i, j, int);
-}
-
-static xbt_matrix_t diff_build_LCS(xbt_dynar_t da, xbt_dynar_t db)
-{
- xbt_matrix_t C =
- xbt_matrix_new(xbt_dynar_length(da), xbt_dynar_length(db),
- sizeof(int), NULL);
- int i, j;
-
- /* Compute the LCS */
- /*
- function LCSLength(X[1..m], Y[1..n])
- C = array(0..m, 0..n)
- for i := 0..m
- C[i,0] = 0
- for j := 1..n
- C[0,j] = 0
- for i := 1..m
- for j := 1..n
- if X[i] = Y[j]
- C[i,j] := C[i-1,j-1] + 1
- else:
- C[i,j] := max(C[i,j-1], C[i-1,j])
- return C[m,n]
- */
- for (i = 0; i < (int)xbt_dynar_length(da); i++)
- for (j = 0; j < (int)xbt_dynar_length(db); j++) {
-
- if (!strcmp(xbt_dynar_get_as(da, i, char *),
- xbt_dynar_get_as(db, j, char *)))
- *((int *) xbt_matrix_get_ptr(C, i, j)) = diff_get(C, i - 1, j - 1) + 1;
- else
- *((int *) xbt_matrix_get_ptr(C, i, j)) = max(diff_get(C, i, j - 1),
- diff_get(C, i - 1, j));
- }
- return C;
-}
-
-static void diff_build_diff(xbt_dynar_t res,
- xbt_matrix_t C,
- xbt_dynar_t da, xbt_dynar_t db, int i, int j)
-{
- char *topush;
- /* Construct the diff
- function printDiff(C[0..m,0..n], X[1..m], Y[1..n], i, j)
- if i > 0 and j > 0 and X[i] = Y[j]
- printDiff(C, X, Y, i-1, j-1)
- print " " + X[i]
- else
- if j > 0 and (i = 0 or C[i,j-1] >= C[i-1,j])
- printDiff(C, X, Y, i, j-1)
- print "+ " + Y[j]
- else if i > 0 and (j = 0 or C[i,j-1] < C[i-1,j])
- printDiff(C, X, Y, i-1, j)
- print "- " + X[i]
- */
-
- if (i >= 0 && j >= 0 && !strcmp(xbt_dynar_get_as(da, i, char *),
- xbt_dynar_get_as(db, j, char *))) {
- diff_build_diff(res, C, da, db, i - 1, j - 1);
- topush = bprintf(" %s", xbt_dynar_get_as(da, i, char *));
- xbt_dynar_push(res, &topush);
- } else if (j >= 0 &&
- (i == -1 || diff_get(C, i, j - 1) >= diff_get(C, i - 1, j))) {
- diff_build_diff(res, C, da, db, i, j - 1);
- topush = bprintf("+ %s", xbt_dynar_get_as(db, j, char *));
- xbt_dynar_push(res, &topush);
- } else if (i >= 0 &&
- (j == -1 || diff_get(C, i, j - 1) < diff_get(C, i - 1, j))) {
- diff_build_diff(res, C, da, db, i - 1, j);
- topush = bprintf("- %s", xbt_dynar_get_as(da, i, char *));
- xbt_dynar_push(res, &topush);
- }
-}
-
-/** @brief Compute the unified diff of two strings */
-char *xbt_str_diff(const char *a, const char *b)
+long int xbt_str_parse_int(const char* str, const char* error_msg)
{
- xbt_dynar_t da = xbt_str_split(a, "\n");
- xbt_dynar_t db = xbt_str_split(b, "\n");
- xbt_matrix_t C;
- xbt_dynar_t diff;
- char *res;
- size_t len;
-
- /* Clean empty lines at the end of da and db */
- len = strlen(a);
- if (len > 0 && a[len - 1] == '\n') {
- char *str;
- xbt_dynar_pop(da, &str);
- free(str);
- }
- len = strlen(b);
- if (len > 0 && b[len - 1] == '\n') {
- char *str;
- xbt_dynar_pop(db, &str);
- free(str);
- }
-
- C = diff_build_LCS(da, db);
- diff = xbt_dynar_new(sizeof(char *), &xbt_free_ref);
-
- diff_build_diff(diff, C, da, db, xbt_dynar_length(da) - 1,
- xbt_dynar_length(db) - 1);
- res = xbt_str_join(diff, "\n");
+ char *endptr;
+ if (str == NULL || str[0] == '\0')
+ THROWF(arg_error, 0, error_msg, str);
- xbt_dynar_free(&da);
- xbt_dynar_free(&db);
- xbt_dynar_free(&diff);
- xbt_matrix_free(C);
+ long int res = strtol(str, &endptr, 10);
+ if (endptr[0] != '\0')
+ THROWF(arg_error, 0, error_msg, str);
return res;
}
-
-
-/** @brief creates a new string containing what can be read on a fd
+/** @brief Parse a double out of a string, or raise an error
*
+ * The #str is passed as argument to your #error_msg, as follows:
+ * THROWF(arg_error, 0, error_msg, str);
*/
-char *xbt_str_from_file(FILE * file)
+double xbt_str_parse_double(const char* str, const char* error_msg)
{
- xbt_strbuff_t buff = xbt_strbuff_new();
- char *res;
- char bread[1024];
- memset(bread, 0, 1024);
+ char *endptr;
+ if (str == NULL || str[0] == '\0')
+ THROWF(arg_error, 0, error_msg, str);
- while (!feof(file)) {
- int got = fread(bread, 1, 1023, file);
- bread[got] = '\0';
- xbt_strbuff_append(buff, bread);
- }
+ double res = strtod(str, &endptr);
+ if (endptr[0] != '\0')
+ THROWF(arg_error, 0, error_msg, str);
- res = buff->data;
- xbt_strbuff_free_container(buff);
return res;
}
#ifdef SIMGRID_TEST
#include "xbt/str.h"
+XBT_TEST_SUITE("xbt_str", "String Handling");
+
#define mytest(name, input, expected) \
xbt_test_add(name); \
d=xbt_str_split_quoted(input); \
input,s,expected);\
free(s); \
xbt_dynar_free(&d);
-
-XBT_TEST_SUITE("xbt_str", "String Handling");
XBT_TEST_UNIT("xbt_str_split_quoted", test_split_quoted, "test the function xbt_str_split_quoted")
{
xbt_dynar_t d;
mytest_str("Basic test", "toto##tutu", "##", "totoXXXtutu");
}
-/* Last args are format string and parameters for xbt_test_add */
-#define mytest_diff(s1, s2, diff, ...) \
+#define test_parse_error(function, name, variable, str) \
do { \
- char *mytest_diff_res; \
- xbt_test_add(__VA_ARGS__); \
- mytest_diff_res = xbt_str_diff(s1, s2); \
- xbt_test_assert(!strcmp(mytest_diff_res, diff), \
- "Wrong output:\n--- got:\n%s\n--- expected:\n%s\n---", \
- mytest_diff_res, diff); \
- free(mytest_diff_res); \
+ xbt_test_add(name); \
+ TRY { \
+ variable = function(str, "Parse error"); \
+ xbt_test_fail("The test '%s' did not detect the problem",name ); \
+ } CATCH(e) { \
+ if (e.category != arg_error) { \
+ xbt_test_exception(e); \
+ } else { \
+ xbt_ex_free(e); \
+ } \
+ } \
+ } while (0)
+#define test_parse_ok(function, name, variable, str, value) \
+ do { \
+ xbt_test_add(name); \
+ TRY { \
+ variable = function(str, "Parse error"); \
+ } CATCH(e) { \
+ xbt_test_exception(e); \
+ } \
+ xbt_test_assert(variable == value, "Fail to parse '%s'", str); \
} while (0)
-XBT_TEST_UNIT("xbt_str_diff", test_diff, "test the function xbt_str_diff")
+XBT_TEST_UNIT("xbt_str_parse", test_parse, "Test the parsing functions")
{
- unsigned i;
-
- /* Trivial cases */
- mytest_diff("a", "a", " a", "1 word, no difference");
- mytest_diff("a", "A", "- a\n+ A", "1 word, different");
- mytest_diff("a\n", "a\n", " a", "1 line, no difference");
- mytest_diff("a\n", "A\n", "- a\n+ A", "1 line, different");
-
- /* Empty strings */
- mytest_diff("", "", "", "empty strings");
- mytest_diff("", "a", "+ a", "1 word, added");
- mytest_diff("a", "", "- a", "1 word, removed");
- mytest_diff("", "a\n", "+ a", "1 line, added");
- mytest_diff("a\n", "", "- a", "1 line, removed");
- mytest_diff("", "a\nb\nc\n", "+ a\n+ b\n+ c", "4 lines, all added");
- mytest_diff("a\nb\nc\n", "", "- a\n- b\n- c", "4 lines, all removed");
-
- /* Empty lines */
- mytest_diff("\n", "\n", " ", "empty lines");
- mytest_diff("", "\n", "+ ", "empty line, added");
- mytest_diff("\n", "", "- ", "empty line, removed");
-
- mytest_diff("a", "\na", "+ \n a", "empty line added before word");
- mytest_diff("a", "a\n\n", " a\n+ ", "empty line added after word");
- mytest_diff("\na", "a", "- \n a", "empty line removed before word");
- mytest_diff("a\n\n", "a", " a\n- ", "empty line removed after word");
-
- mytest_diff("a\n", "\na\n", "+ \n a", "empty line added before line");
- mytest_diff("a\n", "a\n\n", " a\n+ ", "empty line added after line");
- mytest_diff("\na\n", "a\n", "- \n a", "empty line removed before line");
- mytest_diff("a\n\n", "a\n", " a\n- ", "empty line removed after line");
-
- mytest_diff("a\nb\nc\nd\n", "\na\nb\nc\nd\n", "+ \n a\n b\n c\n d",
- "empty line added before 4 lines");
- mytest_diff("a\nb\nc\nd\n", "a\nb\nc\nd\n\n", " a\n b\n c\n d\n+ ",
- "empty line added after 4 lines");
- mytest_diff("\na\nb\nc\nd\n", "a\nb\nc\nd\n", "- \n a\n b\n c\n d",
- "empty line removed before 4 lines");
- mytest_diff("a\nb\nc\nd\n\n", "a\nb\nc\nd\n", " a\n b\n c\n d\n- ",
- "empty line removed after 4 lines");
-
- /* Missing newline at the end of one of the strings */
- mytest_diff("a\n", "a", " a", "1 line, 1 word, no difference");
- mytest_diff("a", "a\n", " a", "1 word, 1 line, no difference");
- mytest_diff("a\n", "A", "- a\n+ A", "1 line, 1 word, different");
- mytest_diff("a", "A\n", "- a\n+ A", "1 word, 1 line, different");
-
- mytest_diff("a\nb\nc\nd", "a\nb\nc\nd\n", " a\n b\n c\n d",
- "4 lines, no newline on first");
- mytest_diff("a\nb\nc\nd\n", "a\nb\nc\nd", " a\n b\n c\n d",
- "4 lines, no newline on second");
-
- /* Four lines, all combinations of differences */
- for (i = 0 ; i < (1U << 4) ; i++) {
- char d2[4 + 1];
- char s2[4 * 2 + 1];
- char res[4 * 8 + 1];
- char *pd = d2;
- char *ps = s2;
- char *pr = res;
- unsigned j = 0;
- while (j < 4) {
- unsigned k;
- for (/* j */ ; j < 4 && !(i & (1U << j)) ; j++) {
- *pd++ = "abcd"[j];
- ps += sprintf(ps, "%c\n", "abcd"[j]);
- pr += sprintf(pr, " %c\n", "abcd"[j]);
- }
- for (k = j ; k < 4 && (i & (1U << k)) ; k++) {
- *pd++ = "ABCD"[k];
- ps += sprintf(ps, "%c\n", "ABCD"[k]);
- pr += sprintf(pr, "- %c\n", "abcd"[k]);
- }
- for (/* j */ ; j < k ; j++) {
- pr += sprintf(pr, "+ %c\n", "ABCD"[j]);
- }
- }
- *pd = '\0';
- *--pr = '\0'; /* strip last '\n' from expected result */
- mytest_diff("a\nb\nc\nd\n", s2, res,
- "compare (abcd) with changed (%s)", d2);
- }
-
- /* Subsets of four lines, do not test for empty subset */
- for (i = 1 ; i < (1U << 4) ; i++) {
- char d2[4 + 1];
- char s2[4 * 2 + 1];
- char res[4 * 8 + 1];
- char *pd = d2;
- char *ps = s2;
- char *pr = res;
- unsigned j = 0;
- while (j < 4) {
- for (/* j */ ; j < 4 && (i & (1U << j)) ; j++) {
- *pd++ = "abcd"[j];
- ps += sprintf(ps, "%c\n", "abcd"[j]);
- pr += sprintf(pr, " %c\n", "abcd"[j]);
- }
- for (/* j */; j < 4 && !(i & (1U << j)) ; j++) {
- pr += sprintf(pr, "- %c\n", "abcd"[j]);
- }
- }
- *pd = '\0';
- *--pr = '\0'; /* strip last '\n' from expected result */
- mytest_diff("a\nb\nc\nd\n", s2, res,
- "compare (abcd) with subset (%s)", d2);
-
- for (pr = res ; *pr != '\0' ; pr++)
- if (*pr == '-')
- *pr = '+';
- mytest_diff(s2, "a\nb\nc\nd\n", res,
- "compare subset (%s) with (abcd)", d2);
- }
+ xbt_ex_t e;
+ int rint = -9999;
+ test_parse_ok(xbt_str_parse_int, "Parse int", rint, "42", 42);
+ test_parse_ok(xbt_str_parse_int, "Parse 0 as an int", rint, "0", 0);
+ test_parse_ok(xbt_str_parse_int, "Parse -1 as an int", rint, "-1", -1);
+
+ test_parse_error(xbt_str_parse_int, "Parse int + noise", rint, "342 cruft");
+ test_parse_error(xbt_str_parse_int, "Parse NULL as an int", rint, NULL);
+ test_parse_error(xbt_str_parse_int, "Parse '' as an int", rint, "");
+ test_parse_error(xbt_str_parse_int, "Parse cruft as an int", rint, "cruft");
+
+ double rdouble = -9999;
+ test_parse_ok(xbt_str_parse_double, "Parse 42 as a double", rdouble, "42", 42);
+ test_parse_ok(xbt_str_parse_double, "Parse 42.5 as a double", rdouble, "42.5", 42.5);
+ test_parse_ok(xbt_str_parse_double, "Parse 0 as a double", rdouble, "0", 0);
+ test_parse_ok(xbt_str_parse_double, "Parse -1 as a double", rdouble, "-1", -1);
+
+ test_parse_error(xbt_str_parse_double, "Parse double + noise", rdouble, "342 cruft");
+ test_parse_error(xbt_str_parse_double, "Parse NULL as a double", rdouble, NULL);
+ test_parse_error(xbt_str_parse_double, "Parse '' as a double", rdouble, "");
+ test_parse_error(xbt_str_parse_double, "Parse cruft as a double", rdouble, "cruft");
}
#endif /* SIMGRID_TEST */